Some performers reproduce the records, down to the drummer. The records already exist.
I begin there because the question behind this note is usually put to me in the form of an objection: that the rhythm sections on these records are the finest of their era, and that a program which prepares them in advance, rather than recreating them with care, has misunderstood where the music lives. I will take the objection seriously, because its premise is correct.
The premise
The rhythm sections on these records are extraordinary. Bernard Purdie’s half-time shuffle on “Home at Last” and “Babylon Sisters” is a textbook in the proper sense: it is taught. Steve Gadd’s drum solo on “Aja” is one of the most admired drum performances on any popular record of the decade. Chuck Rainey’s bass lines on Aja are models of economy. Jim Keltner plays on “Josie.” I do not dispute a word of any of it, and I have listened to all of it more closely than I would wish to say.
Nor do I dispute that the band cared about these parts to an unusual degree. They are reported to have recorded the same song with several rhythm sections before keeping one. They were exacting about time.
The machine
It is at this point that someone generally mentions Wendel. For those who have not had it mentioned to them: during the making of Gaucho, the band’s engineer, Roger Nichols, built a device, which he named Wendel, that allowed drum sounds to be triggered with a consistency no human drummer would give. It has been cited to me with some frequency, as though it were an argument against the position of this note.
It is the argument for it. The band, at the height of its resources and with the best drummers in the country available to it, concluded that for certain purposes the time was a problem to be solved, and solved it with a machine. Once time is correct, it is correct. The rhythm section is a solved problem. A solved problem does not need to be solved again, nightly, in public. It needs to be reproduced accurately and then left alone, so that attention can be given to the parts of the song that are not solved, and which are, in my view, never finished: the harmony, the voicing and the form.
What is prepared, and how
For each selection, the tempo is mapped from the record at the bar level and then fixed. Drums are placed on a sixteenth-note grid. Shuffles are placed at a triplet ratio, which is the ratio the shuffle implies; I do not use the intermediate swing settings, which imply nothing. Each drum voice has three velocity layers: soft, moderate and accented. There is no humanizing function applied. A randomized error is not a feel. It is an error.
The hi-hat is quantized. This has been remarked upon. It is intentional. The hi-hat marks time, and time should be marked accurately.
Where a record contains a drum solo, as “Aja” does, the passage is represented in the program by the harmony beneath it, which is the part of it that was written. I have explained the same principle, at greater length, in the note on the “Peg” solo.
The bass is a separate case, because it is not prepared. It is played live on the pedals. I regard the bass line as harmonic rather than rhythmic, since it determines what every chord above it means (see the note on the mu major, section 5), and it is therefore performed, not reproduced.
The proportion
Of the 4,313 hours recorded in the arrangement log to August 2026, 173 are recorded against the rhythm section, which is 4.0 percent. I am occasionally told that this is too few. It is the number required. Programming a drum part to a fixed grid from a well-documented record is not difficult. It is also not interesting, which I say without disrespect to the players: the interest was in their playing, and their playing is on the records, where anyone may hear it.
I have considered engaging a drummer. I considered it seriously, in 2011, and again later, and I set it aside for the reason given above and for a second reason, which is that a drummer is a person with views on time, and the program has a view on time already. A second view would have to be accommodated. The program is designed for one performer, and the designing was done with that in mind.
What is left
A listener who comes to the program for the drums will be disappointed, and I would rather say so here than at the door. The drums are correct. They are not the point. One does not attend a recital to hear the page turner, although one would notice immediately if he were absent.
What is left, when time has been solved and set aside, is everything the songwriters wrote: the chords, the way the chords are placed, the order of the sections, and the distance between what the words say and what the harmony says beneath them. That is the material of the program. It is not solved, and I do not expect to solve it. I expect to continue.